The parameterization of vegetation in land surface models plays a major role in the simulation of the surface energy balance and therefore weather and climate prediction. Historically parameters in land surface process models have been assigned based on generalized land surface classifications that do not account for local anomalies in phenology. More recently, however, there have been studies that have incorporated satellite remote sensing data in the parameterization of the vegetation used in land surface models. Satellite data provides better spatial and temporal resolution and so improved sampling of the seasonal variability of critical vegetation parameters such as leaf area index (LAI) and fractional vegetation cover. Our hypothesis is that using these improved remotely-sensed parameters may produce improved land surface simulations.

This study addresses the issue of incorporating satellite remote sensing data into a land data assimilation framework. We use the Global Land Data Assimilation System (GLDAS, http://ldas.gsfc.nasa.gov) currently being developed at NASA’s Goddard Space Flight Center and at NOAA’s National Center for Environmental Prediction. GLDAS currently parameterizes LAI according to a limited set of classes, each of which assigns a seasonally varying LAI climatology. In this study, we incorporate LAI and fractional vegetation cover derived from the Advanced High Resolution Radiometer (AVHRR) and conduct seasonal simulations with and without the AVHRR derived LAI data to diagnose its impact on GLDAS and so possible implications for seasonal weather prediction. We investigate if the improved sampling of the vegetation captured by satellite has made improvements in the prediction of key parameters such as soil moisture and surface temperature. The Community Land Model (CLM) is used in the GLDAS simulations and we evaluate each model runs performance with satellite and in-situ data.